Subtotals and grand totals SQL Pivot - sql

Question

Currently have a script that creates a pivot table with current year values subtraction prior year values.
use devmreports
-- Creates dynamic values for pivot table
DECLARE #cols AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols = STUFF((SELECT ',' + QUOTENAME(month)
from ABR
group by ',' + QUOTENAME(month)
order by datalength(',' + QUOTENAME(month))
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
-- Pivot table for YOY change in booked passengers
set #query
=
'SELECT Region,
CityPair,
Year,
' + #cols + '
FROM
(
SELECT ABR.Region,
ABR.CityPair,
ABR.Year,
ABR.Month,
ABR.Adv_B - ABRP.Adv_B as Total
FROM ABR LEFT OUTER JOIN
ABRP ON
ABR.Month = ABRP.Month AND
ABR.CityPair = ABRP.CityPair) P
PIVOT
(
SUM(Total)
FOR MONTH IN
(
'+#cols+'))as pvt'
execute (#Query)
Current Pivot looks like this:
+------------+----------+----+-----+-----+----+
| Region | CityPair | 8 | 9 | 10 | 11 |
+------------+----------+----+-----+-----+----+
| A | 1 | 16 | 17 | 18 | 7 |
| A | 2 | 17 | -20 | -10 | 1 |
| B | 3 | 5 | 8 | 4 | -3 |
| B | 4 | 21 | 10 | 3 | 2 |
| C | 5 | 15 | -14 | -12 | 1 |
+------------+----------+----+-----+-----+----+
What I would like to have is this:
+-----------------+----------+----+-----+-----+----+
| Region | CityPair | 8 | 9 | 10 | 11 |
+-----------------+----------+----+-----+-----+----+
| A | 1 | 16 | 17 | 18 | 7 |
| A | 2 | 17 | -20 | -10 | 1 |
| A Total | | 33 | -3 | 8 | 8 |
| B | 3 | 5 | 8 | 4 | -3 |
| B | 4 | 21 | 10 | 3 | 2 |
| B Total | | 26 | 18 | 7 | -1 |
| C | 5 | 15 | -14 | -12 | 1 |
| C Total | | 15 | -14 | -12 | 1 |
| Grand Total | | 74 | 1 | 3 | 8 |
+-----------------+----------+----+-----+-----+----+
Any assistance would be greatly appreciated.

Answer 1

My suggestion would be to look at using GROUP BY ROLLUP to get the total rows.
The basic syntax if you were hard-coding the query would be:
select
case
when region is null then 'Grand Total'
when citypair is null then region +' Total'
else region end region,
coalesce(cast(citypair as varchar(10)), '') citypair,
sum([8]) [8],
sum([9]) [9]
from
(
select region, citypair, month, total
from yourtable
) d
pivot
(
sum(total)
for month in ([8], [9])
) piv
GROUP BY rollup(region, citypair);
See SQL Fiddle with Demo. Then to use your dynamic SQL version you could alter the code to use:
-- Creates dynamic values for pivot table
DECLARE #cols AS NVARCHAR(MAX),
#colsRollup AS NVARCHAR(MAX),
#query AS NVARCHAR(MAX)
select #cols = STUFF((SELECT ',' + QUOTENAME(month)
from ABR
group by ',' + QUOTENAME(month)
order by datalength(',' + QUOTENAME(month))
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
select #colsRollup = STUFF((SELECT ', sum(' + QUOTENAME(month)+ ') as '+ QUOTENAME(month)
from ABR
group by ',' + QUOTENAME(month)
order by datalength(',' + QUOTENAME(month))
FOR XML PATH(''), TYPE
).value('.', 'NVARCHAR(MAX)')
,1,1,'')
-- Pivot table for YOY change in booked passengers
set #query
=
'SELECT
case
when region is null then ''Grand Total''
when citypair is null then region +'' Total' '
else region end region,
coalesce(cast(citypair as varchar(10)), '''') citypair,
' + #colsRollup + '
FROM
(
SELECT ABR.Region,
ABR.CityPair,
ABR.Year,
ABR.Month,
ABR.Adv_B - ABRP.Adv_B as Total
FROM ABR LEFT OUTER JOIN
ABRP ON
ABR.Month = ABRP.Month AND
ABR.CityPair = ABRP.CityPair
) P
PIVOT
(
SUM(Total)
FOR MONTH IN ('+#cols+')
)as pvt
GROUP BY rollup(region, citypair);'
execute sp_executesql #Query

Answer 2

That's why I prefer to pivot by hand with aggregates and not by pivot. Here's a query which will show you all sums with totals and grand totals:
select
case
when grouping(Region) = 1 then 'Grand Total'
when grouping(CityPair) = 1 then Region + ' Total'
else Region
end as Region,
isnull(cast(CityPair as nvarchar(max)), '') as CityPair,
sum(case when Month = 8 then Value end) as [8],
sum(case when Month = 9 then Value end) as [9],
sum(case when Month = 10 then Value end) as [10],
sum(case when Month = 11 then Value end) as [11]
from test
group by rollup(Region, CityPair)
sql fiddle demo
Here's dynamic one:
declare #stmt nvarchar(max)
select
#stmt = isnull(#stmt + ',', '') +
'sum(case when Month = ' + cast(Month as nvarchar(max)) +
' then Value end) as [' + cast(Month as nvarchar(max)) + ']'
from (select distinct Month from test) as a
select #stmt = '
select
case
when grouping(Region) = 1 then ''Grand Total''
when grouping(CityPair) = 1 then Region + '' Total''
else Region
end as Region,
isnull(cast(CityPair as nvarchar(max)), ''''), ' + #stmt + '
from test
group by rollup(Region, CityPair)'
exec sp_executesql #stmt = #stmt
sql fiddle demo
As always, tried to make it as readable as possible.
I'm suggesting to use group by rollup instead of with rollup, because last one will be deprecated:
WITH ROLLUP This feature will be removed in a future version of
Microsoft SQL Server. Avoid using this feature in new development
work, and plan to modify applications that currently use this feature.